Abstract
Purpose
Analysis of biological samples in large cohort studies may provide insight into the mechanism of, and risk factors for, disease onset and progression.
Methods
This study describes the methods used to collect biological samples from a large multi-center cohort of childhood cancer survivors and siblings of childhood cancer survivors, and evaluates the predictors of a positive response among these individuals.
Results
Among survivors, female sex, white race/ethnicity, college graduation, never smoking, accessing the health care system in the past two years, and having a second malignant neoplasm were the strongest predictors of returning a sample. Among siblings, a similar demographic profile defined those likely to submit the requested sample.
Conclusion
To reduce selection bias and increase the value of these samples for future analysis, additional phone calls and reminders targeting non-responders are needed to improve response rates among those least likely to respond to a single mailed request.
MEDICAL SUBJECT HEADINGS (MeSH): neoplasm, pediatric, buccal mucosa, cohort study, DNA
INTRODUCTION
Large scale cohort studies play an important role in identifying variability in risk factors that contribute to the onset and progression of disease in populations [1]. These investigations provide critical information about the contributions of personal and lifestyle factors to disease incidence, progression, and mortality. Among high risk populations where individuals are enrolled in a cohort based on the presence of or prior treatment for a specific disease, longitudinal cohort studies also provide information about the contributions of disease and treatment factors to long-term medical, psychological, and social outcomes.
Collection of biological samples in epidemiologic studies notably increases the value of the study. Simple, reliable methods of collecting, mailing, and processing buccal cells now yield adequate quantities of DNA for genomic analyses [2,3]. Rapid advances in DNA collection and processing technology [4] allow researchers to examine the independent, additive, and interactive effect of genetics on disease etiology and long-term outcome in association with data regarding personal and lifestyle factors collected in questionnaires [5].
Genetic information is particularly valuable in our study investigating etiology of subsequent disease, second malignancy, and survival in children treated for childhood cancer. Access to DNA samples from survivors and siblings may help investigators identify genetic risk factors for both complications of therapy and secondary cancers in these survivors and their families. In addition, because childhood cancer survivors are at increased risk for chronic and serious medical conditions [6], and because not all excess risk can be explained by disease, treatment, lifestyle and personal factors, the availability of genetic information may help identify those cancer survivors who are at greatest risk for an adverse health outcome.
This manuscript describes the initial methods used to collect buccal cell blood samples from a large cohort of childhood cancer survivors. We report response rates and characteristics of both cancer survivors and their siblings who did and did not respond to a mailed request for a buccal cell sample, or who did and did not provide a blood sample after confirmation of a second malignancy.
METHODS
Study population
Participants eligible for this analysis were living survivor and sibling members of the Childhood Cancer Survivor Study (CCSS) Cohort. Detailed descriptions of the cohort have been published elsewhere [7–9]. Briefly, the CCSS is a retrospective cohort study of individuals who survived five or more years following a diagnosis of leukemia, central nervous (CNS) system malignancy, Hodgkin lymphoma, non-Hodgkin lymphoma, neuroblastoma, Wilms’ tumor, soft tissue sarcoma, or bone malignancy before the age of 21 years. Survivor participants were diagnosed between 1970 and 1986 at one of 26 participating institutions. Sibling participants were enrolled by selecting a random sample of survivors from among those who reported having a sibling, asking the survivor to identify their closest age sibling, and then contacting that sibling to offer enrollment. Approval for human subjects research was obtained from each institution and informed consent was obtained from participants or, for those younger than 18 at study entry, their parents/guardians, prior to data collection. A separate consent was obtained for buccal cell and for blood collection. Language varied some from center to center based on institutional requirements and individual institutional review board requests, although consent forms all contained essentially the same information. Participants were told that their DNA would be stored and used along with information about them and their treatment exposures to help learn about the impact of genetics on cancer development, progression, and response to treatment. They were also told that their information would not be connected to their name, would always be presented in aggregate, and that the sample might not be used immediately, but stored and used in the future.
Sample Collection
The primary outcome of interest for these analyses were whether or not eligible participants returned buccal cell collection kits or agreed to provide a blood sample. Buccal cell collection kits were mailed to all living participants and siblings who had a current address at the time of contact. The collection kit included a 45 ml sample of Scope® mouthwash, a collection jar, and a sealable plastic bag for the collection jar [10]. The packet also included directions for specimen collection, a letter of explanation, a consent form, and a postage-paid first class mailer with a Styrofoam® insert. Follow-up phone calls specific to biological specimen collection were not provided. Collection kits were resent if requested by the participant.
Blood samples were requested from survivors with confirmed second malignant neoplasms. A letter requesting a blood sample and a consent form were mailed to eligible participants. Two weeks after the letter was mailed, a trained interviewer called the individual to explain reasons for the request and answered questions about the collection process. After receiving the signed consent form, the coordinating center contacted a commercial home phlebotomy provider (Examination Management Services in the United States and Hooper Homes in Canada) to set up an appointment with the survivor and collect the blood specimen.
Independent variables
Diagnosis and treatment information were obtained from the medical record by trained abstractors. For these analyses, primary diagnosis was grouped as leukemia, CNS malignancy, Hodgkin disease, non-Hodgkin lymphoma, Wilms’ tumor, neuroblastoma, soft tissue sarcoma, and bone cancer. Age at diagnosis was summarized in four groups: 1– 5 years, 6–10 years, 11–15 years, and 16–20 years. Treatment for the primary diagnosis was classified as receiving or not receiving surgery, chemotherapy, and radiation, by three specific chemotherapeutic classes (anthracyclines, alkylating agents, epipodophyllotixins), and three radiation sites (brain, chest, and gonads).
Demographic and personal data were obtained from the baseline questionnaire, available for viewing at www.stjude.org/ccss, and from a master database that documented collection kit mailing and return dates. Sex, race/ethnicity, age at sample request, educational attainment, marital status, smoking status, a visit to a health care provider in the two years preceding questionnaire completion, and perceived general health were evaluated as predictors of whether or not a sample was returned. Race/ethnicity was classified as black, Hispanic, white and other, and collapsed for multiple variable analyses into white and non-white. Age at sample request was summarized as 10–19 years, 20–29 years, 30–39 years, 40–49 years and 50 years or older. Educational attainment was classified as college graduate or non-college graduate. Marital status was grouped as divorced/separated/widowed, married or living as married, or never married. Contact with a health care provider in the two years preceding the baseline survey and smoking status were treated as dichotomous outcomes. Participants perceived general health was classified as excellent or good and fair or poor.
The presence of a second malignant neoplasm was also considered as a potential predictor of whether or not an individual returned a buccal cell sample, and was the primary eligibility criteria for whether or not an individual was asked to provide a blood sample. Second malignant neoplasms were confirmed by review of pathology reports provided by the institution where the second cancer was diagnosed. Informed consent for release of the pathology report was requested and obtained from individuals who reported a second malignant neoplasm prior to requesting the pathology report.
Statistical methods
The numbers and percentages of individuals who returned and did not return biological samples were calculated by demographic and treatment characteristics and compared in univariate models with Chi-square statistics, or when expected cell sizes were less than five, with Fisher’s exact tests. Multiple variable logistic regression models were used to calculate the relative odds of returning a biological sample by diagnosis, treatment modality, demographics, smoking status, and second malignant neoplasm status among survivors, and by demographic and personal factors among the sibling comparison group. Analyses were done using SAS version 9.2 (Cary, N.C.)
RESULTS
Buccal cell samples
Between April 1996 and December 2007, buccal cell samples were requested from 12,495 survivor and 3154 sibling members of the CCSS cohort. Excluded from analyses were 449 (3.6%) individuals who could not be located, 1305 (10.4%) who did not have an available medical record, and because we used second malignant neoplasm as a predictor of response rate in our models, 385 (3.1%) who could not be confirmed to be free of a second malignant neoplasm. Samples were received from 5,537 (53.5%) of the 10,356 eligible survivors and 1230 (39.0%) of the 3154 siblings. The median time from completion of the baseline questionnaire to the sample request was 5.1 (range 0–11.4) years for survivors and 3.3 (range 0–10.4) years for siblings.
The characteristics of survivors who did and did not return buccal cell samples are shown in Table I by diagnosis and treatment, personal and demographic factors, and by second malignant neoplasm status. Initial univariate examination indicated small differences between the two groups by diagnosis, age at diagnosis, location of radiation exposure, current age, marital status, and perceived general health. However, these differences were small and either did not persist or did not improve the model fit in multiple variable models (Table II). In both univariate and multiple variable models, female sex, white race/ethnicity, college graduation, never-smoking status, recent contact with a health care provider, and having a confirmed second malignant neoplasm were associated with returning a buccal cell sample. Females were 1.2 (95% Confidence Interval (CI): 1.1, 1.3) times more likely to return a sample than were males. Individuals who reported their race/ethnicity as white were 1.9 (95% CI: 1.7, 2.2) times more likely to return a sample than those who reported their race/ethnicity as non-white, and participants with college degrees were 1.2 (95% CI: 1.1, 1.4) times more likely than those without college degrees to return a buccal cell sample. Never-smokers were 20% (95% CI: 10, 30%) more likely than smokers to return a sample. Individuals who had recent contact with a health care provider were 30% (95% CI: 10, 50%) more likely than those who had not had recent contact with a health care provider to return a sample, and those who had experienced a second malignant neoplasm were 40% (95% CI: 20, 70%) more likely than those who had not had a second malignancy to return a buccal cell sample.
Table I.
Characteristics of Survivors and Siblings Who Did and Did Not Return a Buccal Cell Sample by Demographic and Treatment Information
| Survivors | Siblings | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Returned sample (N=5537) |
Refused or non- responder (N=4819) |
p- value$ |
Returned sample (N=1230) |
Refused or non- responder (N=1924) |
p- value$ |
|||||
| N | % | N | % | N | % | N | % | |||
| Gender | ||||||||||
| Male | 2776 | (50.1) | 2633 | (54.6) | <0.001 | 556 | (45.2) | 977 | (50.8) | <0.01 |
| Female | 2761 | (49.9) | 2186 | (45.4) | 674 | (54.8) | 947 | (49.2) | ||
| Race | ||||||||||
| Black | 125 | (2.3) | 222 | (4.6) | 17 | (1.4) | 60 | (3.1) | ||
| Hispanic | 186 | (3.4) | 286 | (5.9) | 31 | (2.5) | 83 | (4.3) | ||
| White | 5080 | (91.6) | 4113 | (85.4) | <0.001 | 1096 | (89.1) | 1656 | (86.1) | <0.01 |
| Other | 128 | (2.3) | 177 | (3.7) | 34 | (2.8) | 59 | (3.1) | ||
| Not reported | 18 | (0.3) | 21 | (0.4) | 52 | (4.2) | 66 | (3.4) | ||
|
Age at buccal request (years) |
||||||||||
| 10 – 19 | 832 | (15.0) | 637 | (13.2) | <0.01 | 5 | (0.4) | 8 | (0.4) | <0.001 |
| 20 – 29 | 2345 | (42.4) | 2185 | (45.3) | 224 | (18.2) | 276 | (14.4) | ||
| 30 – 39 | 1837 | (33.2) | 1605 | (33.3) | 404 | (32.8) | 748 | (38.9) | ||
| 40 – 49 | 511 | (9.2) | 381 | (7.9) | 395 | (32.1) | 643 | (33.4) | ||
| 50 + | 12 | (0.2) | 11 | (0.2) | 183 | (14.9) | 230 | (12.0) | ||
| Educational attainment | 19 | (1.5) | 19 | (1.0) | ||||||
| Less than college | 3911 | (70.6) | 3583 | (74.4) | <0.001 | |||||
| College graduate | 1389 | (25.1) | 958 | (19.9) | 773 | (62.9) | 1367 | (71.1) | <0.001 | |
| Not reported | 237 | (4.3) | 278 | (5.8) | 404 | (32.9) | 462 | (24.0) | ||
| Marital status | 53 | (4.3) | 95 | (4.9) | ||||||
| Divorced/Separated/Wido wed |
292 | (5.3) | 275 | (5.7) | <0.001 | |||||
| Married/Living as married | 1684 | (30.4) | 1295 | (26.9) | 77 | (6.3) | 140 | (7.3) | 0.07 | |
| Never married | 3329 | (60.1) | 3036 | (63.0) | 530 | (43.1) | 762 | (39.6) | ||
| Not reported | 232 | (4.2) | 213 | (4.4) | 563 | (45.8) | 952 | (49.5) | ||
| Smoking status | 60 | (4.9) | 70 | (3.6) | ||||||
| Ever | 1143 | (20.6) | 1136 | (23.6) | <0.001 | |||||
| Never | 4175 | (75.4) | 3517 | (73.0) | 351 | (28.5) | 639 | (33.2) | 0.01 | |
| Not reported | 219 | (4.0) | 166 | (3.4) | 816 | (66.3) | 1216 | (63.2) | ||
|
Accessed health care in the last 2 years |
63 | (5.1) | 69 | (3.6) | ||||||
| No | 491 | (8.9) | 570 | (11.8) | <0.001 | |||||
| Yes | 5045 | (91.1) | 4248 | (88.2) | 187 | (15.2) | 385 | (20.0) | <0.001 | |
| Not reported | 1 | (0.0) | 1 | (0.0) | 1043 | (84.8) | 1539 | (80.0) | ||
| General health | ||||||||||
| Excellent/Good | 5078 | (91.7) | 4368 | (90.6) | 0.08 | 1169 | (95.0) | 1824 | (94.8) | 0.51 |
| Fair/Poor | 415 | (7.5) | 405 | (8.4) | 47 | (3.8) | 83 | (4.3) | ||
| Not reported | 44 | (0.8) | 46 | (1.0) | 14 | (1.1) | 17 | (0.9) | ||
| Cancer Type | ||||||||||
| Leukemia | 1846 | (33.3) | 1712 | (35.5) | <0.001 | |||||
| Central nervous system | 695 | (12.6) | 578 | (12.0) | ||||||
| Hodgkin disease | 768 | (13.9) | 523 | (10.9) | ||||||
| Non-Hodgkin lymphoma | 397 | (7.2) | 387 | (8.0) | ||||||
| Kidney (Wilms) | 540 | (9.8) | 425 | (8.8) | ||||||
| Neuroblastoma | 392 | (7.1) | 343 | (7.1) | ||||||
| Soft tissue sarcoma | 471 | (8.5) | 431 | (8.9) | ||||||
| Bone cancer | 428 | (7.7) | 420 | (8.7) | ||||||
| Age at diagnosis (years) | ||||||||||
| 0 – 5 | 2595 | (46.9) | 2370 | (49.2) | <0.01 | |||||
| 6 – 10 | 1038 | (18.8) | 952 | (19.8) | ||||||
| 11 – 15 | 1180 | (21.3) | 937 | (19.4) | ||||||
| 16 – 20 | 724 | (13.1) | 560 | (11.6) | ||||||
|
Treatment for primary diagnosis* |
||||||||||
| Surgery | 4464 | (80.6) | 3839 | (79.7) | 0.20 | |||||
| Chemotherapy | 4408 | (79.6) | 3884 | (80.6) | 0.21 | |||||
| Anthracyclines | 2160 | (39.0) | 1963 | (40.7) | 0.08 | |||||
| Alkylating agents | 2842 | (51.3) | 2457 | (51.0) | 0.72 | |||||
| Epipodophyllotoxins | 450 | (8.1) | 351 | (7.3) | 0.11 | |||||
| Radiation | 3650 | (65.9) | 3122 | (64.8) | 0.23 | |||||
| Cranial | 1718 | (31.0) | 1537 | (31.9) | 0.14 | |||||
| Chest | 1139 | (20.6) | 863 | (17.9) | <0.01 | |||||
| Gonads | 609 | (11.0) | 447 | (9.2) | <0.01 | |||||
| Second malignancy# | ||||||||||
| None | 4934 | (89.1) | 4466 | (92.7) | <0.001 | |||||
| Non-Malignant Skin Cancer |
304 | (5.5) | 177 | (3.7) | ||||||
| Breast | 79 | (1.4) | 37 | (0.8) | ||||||
| Central Nervous System | 25 | (0.5) | 7 | (0.2) | ||||||
| Leukemia/ Lymphoma | 26 | (0.5) | 16 | (0.3) | ||||||
| Meningioma | 5 | (0.1) | 2 | (0.0) | ||||||
| Thyroid | 56 | (1.0) | 45 | (0.9) | ||||||
| Other | 107 | (1.9) | 69 | (1.4) | ||||||
Categories are not mutually exclusive
First second malignancy
P-value does not account for group who did not report the variable, all P-values are two-sided
Table II.
Percentages of Childhood Cancer Survivors and Siblings Who Returned a Buccal Cell Sample With Relative Odds and 95% Confidence Intervals From Multiple Variable Models
| Childhood Cancer Survivors |
Siblings | |||||
|---|---|---|---|---|---|---|
| Percent returned |
Odds ratio |
95% CI | Percent returned |
Odds ratio |
95% CI | |
| Gender | ||||||
| Male | 51.5 | 1.0 | 35.6 | 1.0 | ||
| Female | 56.1 | 1.2 | 1.1, 1.3 | 41.4 | 1.2 | 1.1, 1.4 |
| Race/ethnicity | ||||||
| non-white | 38.7 | 1.0 | 27.6 | 1.0 | ||
| white | 55.5 | 1.9 | 1.7, 2.2 | 39.7 | 1.7 | 1.2, 2.2 |
| Highest Education level | ||||||
| College graduate | 59.1 | 1.2 | 1.1, 1.4 | 46.6 | 1.5 | 1.3, 1.8 |
| Non-college graduate | 51.9 | 1.0 | 35.2 | 1.0 | ||
| Never Smoked | ||||||
| No | 50.3 | 1.0 | 34.6 | 1.0 | ||
| Yes | 54.7 | 2.0 | 1.1, 1.4 | 40.5 | 1.3 | 1.1, 1.5 |
| Accessed health care in last 2 years | ||||||
| Yes | 54.6 | 1.3 | 1.1, 1.5 | 40.0 | 1.3 | 1.1, 1.6 |
| No | 46.1 | 1.0 | 32.2 | 1.0 | ||
| Second malignant neoplasm | ||||||
| No | 52.6 | 1.4 | Not Applicable | |||
| Yes | 64.2 | 1.0 | 1.3, 1.7 | |||
CI=confidence interval
The characteristics of siblings who returned and did not return a buccal cell sample are shown in Table I. Again, female gender, race/ethnicity, educational attainment, smoking status and contact with a health care provided were significant predictors of whether or not an individual returned a biological sample (Table II). Female siblings were 1.2 (95% CI: 1.1, 1.4) times likely than male siblings to return a sample. Like survivors, siblings who reported their race/ethnicity as white were 1.7 (95% CI: 1.2, 2.2) times more likely than those who reported their race as non-white to return a sample, and those with college degrees were 1.5 (95% CI: 1.3, 1.8) times more likely than those without college degrees to return a buccal cell sample. Finally, non-smokers were 1.3 (95% CI: 1.1,1.5) times more likely than ever smokers, and those who had seen a health care provider in the two years preceding the baseline survey were 1.3 (95% CI: 1.1,1.6) times more likely than those who had not to return a biological sample.
Blood samples
From August 2001 to July 2007, blood samples were requested from 434 survivors who had a confirmed second malignancy and an available medical record, a median of 7.7 (range 2.9–13.4) years from completion of the baseline questionnaire. Samples were returned by 230 (53%) survivors. Blood sample return rates among survivors with a second malignancy differed only by race/ethnicity and treatment history. Responders were more likely to be white then non-white (Odds Ratio (OR): 2.8; 95% CI: 1.3, 6.2), and those who received anthracyclines were less likely than those who had not to return a blood sample (OR: 0.6; 95% CI: 0.4, 0.9).
DISCUSSION
The results of these analyses indicate that just over half of survivors and slightly more than one-third of sibling members of an ongoing cohort of long-term survivors of pediatric and adolescent cancer provided a buccal cell sample for DNA analysis via mail. Female non-smoking college graduates who report their race as white, and who have regular contact with their health care provider have the highest response rates among both survivors and siblings. Survivors who experience a second malignant neoplasm are also more likely to provide a sample. Among survivors with a second malignant neoplasm who are asked to provide a blood sample, white race is again predictor of participation.
Our overall response rate (53.5%) was lower than those reported in two large community- and profession-based cohort studies, and similar to those reported in a large population-based cohort study. LeMarchand et al. [11] successfully collected buccal cell samples from two-thirds of a Hawaiian cohort, and Hansen et al. [2] successfully collected saliva samples from 72% of persons in the Danish Nurses cohort study. Zheng et al. [12] reported a 49.3% participation rate for buccal cell collection in a population based cohort of urban Chinese women. Specific follow-up reminders were provided in both the Hawaiian and Dutch cohorts, but not in our study or in the Shanghai Women’s Health Study. Routine phone contact after mailing a request may enhance participation, increase the available samples, and perhaps decrease potential bias when genetic analyses are done using samples from the study population.
The characteristics of participants who were most likely to provide a buccal cell sample in our study are similar to study responder characteristics in other studies where a biological specimen was requested. Engel et al. [13] reported that white, educated farmers with a history of melanoma were more likely to return a buccal cell sample than those who reported their race as non-white, who had less than a college degree, and who did not have a history of melanoma. Cozier [14] reported a buccal cell sample return participation rate of 37% among members of the Black Women’s Health Study who had been diagnosed with cancer, and Garoufalis et al. [15] reported that individuals with more severe myopia were more likely than those with less severe myopia to enroll in an observational trial when blood specimens for DNA analysis were collected. Like these investigators, we found that participation was better among white, educated individuals, and that responders were more likely to have a severe outcome, incident second cancers, than were non-responders.
The CCSS is a resource for research [9], and as such, is committed to providing quality, representative data and biological materials for future analyses. A representative sample from this large cancer survivor population is essential to allow for an unbiased assessment of the complex relationships between genes, the environment, and disease. [16] A sample that includes survivors with diverse personal and demographic characteristics will optimize researchers’ abilities to discover important genes that infer risk for either cancer or a poor outcome, and eventually will allow for tailored treatments and prevention plans. If a specific sub-group of the survivor population is not represented by their genetic material, findings that inform future health care and prevention measures will not be able to be generalized to them.
Investigators have recently begun a project to supplement and enhance the availability of the DNA samples. All living survivors and siblings in both the original (children diagnosed between 1970 and 1986) and the expansion cohorts (children diagnosed between 1987 and 1999) are being asked to provide additional saliva sample. Original cohort participants who return the 2007 follow-up questionnaire, and expanded cohort participants who return their baseline questionnaire, are mailed an Oragene® saliva collection kit with a consent form, an instruction sheet, a plastic biohazard bag to place the sample in, and a return pre-paid padded envelope. If the sample has not been returned with in three weeks, up to ten phone call attempts are made to speak with the participant to encourage them to return the sample. Phone call attempts are made at different times of the day and on different days of the week. Up to three verbal messages are left if study personnel are confident that they have the correct contact information for a particular survivor. A second collection kit is mailed if the participant has not received or has misplaced the original kit. The saliva collection kit and phone call attempts are repeated after six months for initial non-responders. Using this multiple contact technique, with a phone call, a potential additional mailing, and a re-contact attempt six months after the initial mailing should improve response rates. Preliminary phone interactions between study participants and research personnel have primarily involved answering questions and/or clarifying information on the consent form (Figure 1). Qualitative responses in these conversations have been positive.
Figure 1.
Frequently Asked Questions Regarding Oragene® Saliva Collection and Usual Research Staff Responses
Acknowledgements
This work was supported by grant CA 55727 (L.L. Robison, Principal Investigator), National Cancer Institute, Bethesda, MD, with additional support provided to St. Jude Children’s Research Hospital by ALSAC.
LIST OF ABBREVIATIONS
- DNA
Deoxyribonucleic Acid
- CCSS
Childhood Cancer Survivor Study
- CNS
Central Nervous System
References
- 1.Calle EE, Rodriguez C, Jacobs EJ, et al. The American Cancer Society Cancer Prevention Study II Nutrition Cohort: rationale, study design, and baseline characteristics. Cancer. 2002;94:2490–2501. doi: 10.1002/cncr.101970. [DOI] [PubMed] [Google Scholar]
- 2.Hansen TV, Simonsen MK, Nielsen FC, et al. Collection of blood, saliva, and buccal cell samples in a pilot study on the Danish nurse cohort: comparison of the response rate and quality of genomic DNA. Cancer Epidemiol Biomarkers Prev. 2007;16:2072–2076. doi: 10.1158/1055-9965.EPI-07-0611. [DOI] [PubMed] [Google Scholar]
- 3.Harty LC, Shields PG, Winn DM, et al. Self-collection of oral epithelial cell DNA under instruction from epidemiologic interviewers. Am J Epidemiol. 2000;151:199–205. doi: 10.1093/oxfordjournals.aje.a010188. [DOI] [PubMed] [Google Scholar]
- 4.Zheng S, Ma X, Buffler PA, et al. Whole genome amplification increases the efficiency and validity of buccal cell genotyping in pediatric populations. Cancer Epidemiol Biomarkers Prev. 2001;10:697–700. [PubMed] [Google Scholar]
- 5.Hayes RB, Sigurdson A, Moore L, et al. Methods for etiologic and early marker investigations in the PLCO trial. Mutat Res. 2005;592:147–154. doi: 10.1016/j.mrfmmm.2005.06.013. [DOI] [PubMed] [Google Scholar]
- 6.Oeffinger KC, Mertens AC, Sklar CA, et al. Chronic health conditions in adult survivors of childhood cancer. N Engl J Med. 2006;355:1572–1582. doi: 10.1056/NEJMsa060185. [DOI] [PubMed] [Google Scholar]
- 7.Robison LL, Mertens AC, Boice JD, et al. Study design and cohort characteristics of the Childhood Cancer Survivor Study: a multi-institutional collaborative project. Med Pediatr Oncol. 2002;38:229–239. doi: 10.1002/mpo.1316. [DOI] [PubMed] [Google Scholar]
- 8.Leisenring WM, Mertens AC, Armstrong GT, et al. Pediatric cancer survivorship research: experience of the Childhood Cancer Survivor Study. J Clin Oncol. 2009;27:2319–2327. doi: 10.1200/JCO.2008.21.1813. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Robison LL, Armstrong GT, Boice JD, et al. The Childhood Cancer Survivor Study: a National Cancer Institute-supported resource for outcome and intervention research. J Clin Oncol. 2009;27:2308–2318. doi: 10.1200/JCO.2009.22.3339. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Heath EM, Morken NW, Campbell KA, et al. Use of buccal cells collected in mouthwash as a source of DNA for clinical testing. Arch Pathol Lab Med. 2001;125:127–133. doi: 10.5858/2001-125-0127-UOBCCI. [DOI] [PubMed] [Google Scholar]
- 11.Le Marchand L, Lum-Jones A, Saltzman B, et al. Feasibility of collecting buccal cell DNA by mail in a cohort study. Cancer Epidemiol Biomarkers Prev. 2001;10:701–703. [PubMed] [Google Scholar]
- 12.Zheng W, Chow WH, Yang G, et al. The Shanghai Women's Health Study: rationale, study design, and baseline characteristics. Am J Epidemiol. 2005;162:1123–1131. doi: 10.1093/aje/kwi322. [DOI] [PubMed] [Google Scholar]
- 13.Engel LS, Rothman N, Knott C, et al. Factors associated with refusal to provide a buccal cell sample in the Agricultural Health Study. Cancer Epidemiol Biomarkers Prev. 2002;11:493–496. [PubMed] [Google Scholar]
- 14.Cozier YC, Palmer JR, Rosenberg L. Comparison of methods for collection of DNA samples by mail in the Black Women's Health Study. Ann Epidemiol. 2004;14:117–122. doi: 10.1016/S1047-2797(03)00132-7. [DOI] [PubMed] [Google Scholar]
- 15.Garoufalis P, Chen CY, Dirani M, et al. Methodology and recruitment of probands and their families for the Genes in Myopia (GEM) Study. Ophthalmic Epidemiol. 2005;12:383–392. doi: 10.1080/09286580500281222. [DOI] [PubMed] [Google Scholar]
- 16.Kaiser J. Biobanks. Population databases boom, from Iceland to the U.S. Science. 2002;298:1158–1161. doi: 10.1126/science.298.5596.1158. [DOI] [PubMed] [Google Scholar]